Introduction: Evidence about outcomes in ECMO in trauma-related ARDS from middle-income countries is absent, and mortality scores have never been validated in this environment. We characterised outcomes and complications of poly-traumatised adults treated with ECMO and assessed the accuracy of APACHE II, Injury Severity Score (ISS) and RESP score. Methods: We conducted an ambispective, multicentre cohort across four Argentine ECMO centres (2015-2024). All adults (≥18 yr) cannulated for refractory hypoxaemia after polytrauma were included. Pre-defined variables—demographics, injury pattern, ECMO configuration/duration, anticoagulation strategy and International Society on Thrombosis and Haemostasis (ISTH) bleeding categories—were extracted from electronic charts. Survivors underwent structured telephone follow-up with EQ-5D-3L and Katz Index. Model performance for in-hospital mortality was quantified by AUROC (discrimination) and Brier score + Spiegelhalter z (calibration). Institutional review boards approved the protocol, and survivors provided informed consent. Results: Thirty-two patients (mean age 32 ± 16 yr; 84 % male) received ECMO—31 venovenous, 1 veno-arterial. Median ISS was 34 (IQR 25–50); 63 % had concomitant traumatic brain injury. Time from intubation to cannulation was 4 d (1–6) and ECMO ran for 7 d (3–11). Heparin was administered in 59 %. Device-related events were common: oxygenator/circuit thrombosis 22 %, deep-vein thrombosis 19 %, any bleeding 28 % (major 16 %). Overall in-hospital mortality was 40.6 % (95 % CI 24–59) and occurred early (ECMO day 3 1–7 vs 9.5 6–11.8 in survivors, p = 0.009). Prognostic scores performed only moderately: ISS AUROC 0.63/Brier 0.224, RESP 0.56/0.240, APACHE II 0.54/0.241; pairwise AUROC differences were non-significant (p = 0.75). Nineteen patients survived to discharge; 11 (58 %) were interviewed a median 7 yr later. Median EQ-5D utility was 0.74 (0.65–0.83) and Katz 6 (5–6), yet chronic pain persisted in 55 % and anxiety/depression in 73 %. Conclusions: In this first Latin-American series, ECMO delivered 59 % long-term survival with high functional independence, but bleeding/thrombotic burdens were substantial and existing scores predicted mortality poorly. Trauma-specific risk models and structured post-ICU rehabilitation are urgently needed.
Carini et al. (Sun,) studied this question.